Conventionally, a life cycle of a software application includes at least the following stages: development, quality assurance, and deployment. In addition, updates are often made to software applications after deployment. For example, a patch or an upgrade of the software application may be made via an update.
Typically, a software application may be deployed to multiple platforms or systems having different configurations. Because of the different configurations, these platforms may each have its unique set of requirements for software at different stages of its life cycle. As such, the software application may have to be deployed in different ways depending on the types of platforms.
Because of these complicated considerations, software applications are typically deployed manually in order to accommodate the different requirements of various platforms. For instance, to deploy a software application, all of the content relevant to the application is marked as ready to be deployed in a centralized software repository by tagging, and then exported from the centralized software repository to application servers. Likewise, when there is update of the software application to deploy, one has to manually repeat this process in order to deploy the update to systems that are affected. This is an error-prone, labor-intensive process. If something is incorrectly tagged, either necessary changes or updates do not make it into production, or code plagued with errors that was not meant for production does get into production. To safeguard against the above mishap, a person (such as a system administrator) has to monitor and validate the deployment process at each stage.